On June 17, 1948, Bell Telephone Laboratories filed US Patent 2,524,035, titled "Three-Electrode Circuit Element Utilizing Semiconductive Materials." The inventor of record was John Bardeen and Walter Brattain; a companion filing (US 2,569,347) covered William Shockley's junction transistor variant developed weeks later in a competitive frenzy. The patent describes a small slab of germanium with two phosphor-bronze "cat's whisker" point contacts spaced about 50 microns apart, pressed onto a base electrode. A tiny current injected at one contact (the emitter) modulated a much larger current flowing to the other (the collector). It was an amplifier — but built from a crystal, not a vacuum.

The breakthrough had happened six months earlier, on December 23, 1947, when Bardeen and Brattain demonstrated audio amplification through their crude germanium device to Bell Labs management. Shockley, embarrassed at being absent from the actual discovery, locked himself in a Chicago hotel room over New Year's and worked out the theory of the bipolar junction transistor — a sandwich of doped semiconductor layers that didn't need finicky point contacts at all. His version (filed June 1948, granted 1951) is what actually went into production.

Why it was shocking for 1948: The dominant amplifier was the vacuum tube — hot, fragile, power-hungry, and the size of a thumb. A 1947 telephone exchange ran on thousands of them; ENIAC used 17,468. Everyone knew tubes were the bottleneck for scaling electronics, but nobody had a replacement. Bell Labs' device was solid-state, cool-running, drew milliwatts instead of watts, and could in principle be made arbitrarily small. The press conference on June 30, 1948 buried the announcement on page 46 of the New York Times.

The modern connection is almost too direct to be interesting — except for what the patent didn't anticipate:

• Scale. Bardeen and Brattain's device was hand-assembled and millimeters wide. By 2025, TSMC's 2nm process packed roughly 300 million transistors per square millimeter. The patent's "three-electrode circuit element" now exists in numbers exceeding 10²² — more than the stars in the observable universe, manufactured per year.
• Integration. The patent describes one transistor. Jack Kilby and Robert Noyce's integrated circuit (1958–59) is what made transistors economically interesting, by putting many on one substrate and eliminating the wire-bonding bottleneck.
• The wrong material won. Bardeen and Brattain used germanium because it was easier to refine. Silicon, with its self-passivating oxide, would eventually displace it almost entirely after Atalla's surface-passivation work at Bell Labs in 1959.
• The wrong architecture won. The bipolar transistor of the original patent was eclipsed by the MOSFET — a field-effect device Shockley had actually theorized first but couldn't make work. Today, well over 99% of all transistors manufactured are MOSFETs.

The Nobel Prize came in 1956, shared among all three. By then, Shockley had alienated his colleagues so thoroughly that his Mountain View startup — Shockley Semiconductor — was hemorrhaging staff. Eight of them ("the traitorous eight") left in 1957 to found Fairchild Semiconductor. From Fairchild came Intel, AMD, and the seed of Silicon Valley itself. The patent's most consequential legacy isn't electrical — it's geographic.